中文摘要：

水资源配置具有模糊性、不确定性和多目标性的特点，对其方案进行综合评价需要克服因模糊性和不确定性而引起的各目标权重确定的精确性欠佳问题，以及多目标之间的不可公度性和矛盾性问题。对此设计两个核心步骤来予以解决：①利用证据理论较强的多元信息融合能力，通过对已知信息的挖掘和融合，确定评价指标的权重；②引入格序理论将全序决策拓展为格序决策，选取正、负加权理想解为顶元素和底元素，构造概念格，结合加权Kaufmann距离的概念，以备选方案与正、负加权理想解的综合距离为综合标准来对方案进行优选。以天津市水资源配置方案综合评价为例进行分析，计算及对比分析结果表明，利用该方法对水资源配置方案进行评价客观合理、有效。

英文摘要：

Water resource allocation involves aspects of economy, society, ecology and resources and comprises fuzzification, uncertainty and multi-objectivity. It is therefore important to decrease the error of the weight decision caused by the influence of artificial factors and mine information contained in known data as much as possible to overcome contradiction and incommensurability among multi-objectives during scheme optimization. Here, we undertook two key steps to solve the above problem. First, using evidence theory from the field of data processing and data integration, we determine the weight of each index on the basis of the ideas that the less distance between two pieces of evidence indicates that the support degree of a piece of evidence by others is higher and the evidence is more reliable. Second, in light of the lattice-order theory, through selecting the top and bottom elements as the positive and negative ideal solutions and fusing with obtained weights, we arrived at the concept-lattice. We then selected the optimal from the alternatives by calculating the value of compound distances between alternatives and weighted positive and negative ideal solutions, by means of applying the concept of weighted Kaufmann distance or close-degree. Last, we take comprehensive evaluation schemes for Tianjin, China to verify the validity and feasibility of our evaluation method. We conclude that like the BP neural network model or fuzzy optimum selecting model or fuzzy entropy model, we can find that this method owns preferable practicability, validity and feasibility for the comprehensive evaluation for water resource allocation schemes.